Back to skill

Security audit

prisma-api

Security checks for vulnerabilities and agentic risk

Overview

The skill is coherent for Prisma Access API management, but it gives an agent high-impact curl-based authority to change and push security configuration with limited built-in safeguards.

Review this skill before installing in a production tenant. Use least-privileged or test SCM credentials, avoid exposing the client secret or bearer token in logs, and require explicit human review plus backup or rollback planning before create, update, delete, or candidate push operations.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (11)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill explicitly supports create, update, delete, and push operations against live Prisma Access configuration, but the top-level description does not prominently warn that these actions can modify production security posture and disrupt connectivity. In an automation context, underspecified safety messaging increases the chance of accidental destructive use or unintended policy deployment.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 6)May include surrounding context.

md
description: Interact with the Strata Cloud Manager (SCM) API to manage Prisma Access configurations. Authenticate, query, create, update, and delete configuration objects. Use when automating Prisma Access operations or querying live tenant state.
argument-hint: "[operation] [resource]"
disable-model-invocation: true
allowed-tools: Bash(curl *)
version: 1.1.0
metadata:
  openclaw:

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 65)May include surrounding context.

List / Query Resources

bash
curl -s -X GET "https://api.sase.paloaltonetworks.com/sse/config/v1/{resource}?folder={folder}&limit=200" \
  -H "Authorization: Bearer ${TOKEN}"

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 65)May include surrounding context.

List / Query Resources

bash
curl -s -X GET "https://api.sase.paloaltonetworks.com/sse/config/v1/{resource}?folder={folder}&limit=200" \
  -H "Authorization: Bearer ${TOKEN}"

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 130)May include surrounding context.

List / Query Resources

bash
curl -s -X GET "https://api.sase.paloaltonetworks.com/sse/config/v1/{resource}?folder={folder}&limit=200" \
  -H "Authorization: Bearer ${TOKEN}"

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 137)May include surrounding context.

List / Query Resources

bash
curl -s -X GET "https://api.sase.paloaltonetworks.com/sse/config/v1/{resource}?folder={folder}&limit=200" \
  -H "Authorization: Bearer ${TOKEN}"

External Transmission

Medium
Category
Data Exfiltration
Confidence
84% confidence
Finding

This POST example enables creation of configuration objects on a live security platform without any built-in guardrails in the skill definition beyond advisory text. In practice, write-capable API patterns can be dangerous because malformed or unintended objects may weaken policy, create exposure, or disrupt access when used by an automation agent.

Content

Scanner excerpt · SKILL.md (reported line 93)May include surrounding context.

Create a Resource

bash
curl -s -X POST "https://api.sase.paloaltonetworks.com/sse/config/v1/{resource}?folder={folder}" \
  -H "Authorization: Bearer ${TOKEN}" \
  -H "Content-Type: application/json" \
  -d '{...}'

External Transmission

Medium
Category
Data Exfiltration
Confidence
86% confidence
Finding

The PUT example permits modification of existing configuration objects in a live tenant. In a security-management context, automated updates can silently alter enforcement rules, addresses, services, or profiles, potentially causing outages or reducing protections if parameters are incorrect or attacker-influenced.

Content

Scanner excerpt · SKILL.md (reported line 102)May include surrounding context.

Update a Resource

bash
curl -s -X PUT "https://api.sase.paloaltonetworks.com/sse/config/v1/{resource}/{id}" \
  -H "Authorization: Bearer ${TOKEN}" \
  -H "Content-Type: application/json" \
  -d '{...}'

External Transmission

Medium
Category
Data Exfiltration
Confidence
90% confidence
Finding

The DELETE example supports removal of configuration objects from the tenant, which is inherently destructive. Without enforced confirmation, dependency checks, or rollback guidance, accidental or manipulated deletions could break policy references, disable protections, or interrupt service behavior.

Content

Scanner excerpt · SKILL.md (reported line 111)May include surrounding context.

Delete a Resource

bash
curl -s -X DELETE "https://api.sase.paloaltonetworks.com/sse/config/v1/{resource}/{id}" \
  -H "Authorization: Bearer ${TOKEN}"

External Transmission

Medium
Category
Data Exfiltration
Confidence
93% confidence
Finding

The candidate push operation can promote pending configuration changes toward active deployment, making this the highest-risk action in the skill. Even though the prose advises asking for user review, the skill does not technically enforce that requirement, so an agent could push unsafe or unreviewed changes that affect production security posture and availability.

Content

Scanner excerpt · SKILL.md (reported line 121)May include surrounding context.

bash
# Push candidate config
curl -s -X POST "https://api.sase.paloaltonetworks.com/sse/config/v1/config-versions/candidate:push" \
  -H "Authorization: Bearer ${TOKEN}" \
  -H "Content-Type: application/json" \
  -d '{"folders": ["Prisma Access"]}'

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

The authentication example instructs users to place client credentials and tenant scope values in environment variables and use them directly in curl commands, but it does not warn about secret handling, shell history, process visibility, or accidental logging. While common practice, omission of secret-handling guidance can lead to credential exposure in shared terminals, CI logs, or debugging output.

Content

No source excerpt is available for this finding.